search

INRIA / scikit-learn-mooc

Machine learning in Python with scikit-learn MOOC
https://inria.github.io/scikit-learn-mooc
Creative Commons Attribution 4.0 International
1.12k stars 516 forks source link

FIX Several small fixes #780

Open ArturoAmorQ opened 5 months ago

ArturoAmorQ commented 5 months ago

  1. At the end of the First look at the dataset notebook, under the heading "we made important observations" we mention that linear models can only capture linear interactions but there is no prior reference to linear models.

  2. The word "complex" at the end of the Effect of the sample size in cross-validation notebook may be misleading. Here we don’t mean complexity as in the depth of a decision tree or the degree of a polynomial (which indeed would lead to overfitting), but rather that a more "expressive" model should be used.

  3. In the Solution for Exercise M3.02 the wording refers to the "accuracy of the model" but the model is a KNeighborsRegressor.

  4. The value_counts function in the Classification metrics notebook is raising a warning.

This PR proposes some fixes to address those issues.

ogrisel commented 1 day ago

@ArturoAmorQ I finally answered your feedback in https://github.com/INRIA/scikit-learn-mooc/pull/780#discussion_r1853981184.